MR Imaging of Endometrial Cancer that Occurs After Radiation Therapy for Cervix Cancer

PURPOSE: We wanted to describe the MR imaging findings of endometrial cancer in patients with a history of prior radiation therapy for cervical cancer (ECRT) and we compare them to the MR imaging findings of patients with spontaneously occurring endometrial cancer (SEC). MATERIALS AND METHODS: Twenty-two patients with endometrial cancer that was diagnosed by operation or endometrial biopsy were included in the study. The patients were divided into two groups according to the presence of past RT for cervical cancer: ECRT (n = 4) and SEC (n = 18). The MR images were retrospectively analyzed by consensus of two experienced radiologists. The MR imaging findings were analyzed by the size, shape and signal intensity of the mass, distension of the uterine cavity, the presence of cervical stenosis and the nature of the fluid collection. RESULTS: For the mass shape, all the ECRT lesions were polypoid masses. However, the SEC patients had 5 polypoid masses and 13 wall thickenings. The maximal diameter, signal intensity and enhancement pattern of the masses were not different between the ECRT and SEC patients. The width of the endometrial cavity varied between 3.9 cm in the ECRT patients and 0.4 cm in the SEC patients (p =0.002). All the ECRT patients had cervical stenosis. However, none of the SEC patients had cervical stenosis. CONCLUSION: MR imaging of ECRT patients demonstrated prominent distension of their uterine cavity and cervical stenosis, which may be the result of radiation fibrosis in the uterus.

MRI Findings of Cavernous Hemangioma of the Uterus: A Case Report

Hemangioma is the most common soft tissue tumor in the body. Though it may occur anywhere in the body, hemangioma of the uterus is a very rare tumor. Hemangioma is almost asymptomatic, but it is sometimes clinically important because it can cause massive hemorrhage and this is a life-threatening condition. We report here on the magnetic resonance imaging and pathologic findings of cavernous hemangioma of the uterus in a 32-year-woman with menorrhagia.

The Usefulness of Dynamic MR Imaging for the Evaluation of Cervical Cancer

PURPOSE: We wished to evaluate the diagnostic usefulness of dynamic MRI in assessing tumor visualization and the parametrial invasion of cervical cancer, and we also wished to determine the most adequate enhancing time by comparing the T2-weighted image (T2WI) and enhanced T1-weighted image (Gd-T1WI). MATERIALS AND METHODS: Fifty-three women with histopathologically proven cervical cancer underwent a preoperative MRI. Using a 1.5 T magnet, the fast spin echo axial T2WI without fat saturation was taken; after contrast administration, 20, 40, 60, 90, 120 sec-dynamic MRIs were taken using fast SPGR and spin echo axial Gd-T1WI. Tumor conspicuity and parametrial invasion in each pulse sequence and the most adequate enhancing time for the evaluation of the tumor on dynamic MRI were evaluated prospectively by three radiologists working at three separate sessions. The results were then correlated with the histopathologic findings. RESULTS: The conspicuity of tumor on dynamic MRI (99.4%) and T2WI (95.6%) were better than on Gd-T1WI (89.3%). In the assessment of parametrial invasion of the tumor, the diagnostic accuracy of dynamic MRI, Gd-T1WI and T2WI was 79.9%, 78% and 76.1%, respectively; the highest values were for the dynamic MRI, but there was no statistically significant difference among three pulse sequences. The most adequate enhancing time on dynamic MRI was between 90 seconds and 120 seconds. CONCLUSION: Dynamic MRI is useful for the assessment of tumor visualization of cervical cancer, and the most appropriate scan time on dynamic MRI is between 90 seconds and 120 seconds. For the determination of parametrial invasion, the dynamic MRI revealed a higher diagnostic accuracy than that of T2WI or Gd-T1WI, but the differences were statistically insignificant.

Development of VX-II Carcinoma Model in Rabbit Uterus: Evaluation with MR Imaging and Histopathology

PURPOSE: To develop a VX-II carcinoma model in the rabbit uterus and to describe the MR imaging findings of an experimentally induced VX-II uterine carcinoma along with the histopathologic findings. MATERIALS AND METHODS: 13 New Zealand rabbits were included in this study. Two pieces of tissue suspension (1 mm3x2) of VX-II carcinoma were loaded in an 18 gauge disposable needle and the tissue embedded in the wall of each horn of the rabbits' uterus. We obtained the MR images at 2 weeks in group A (n=5), or at 2 and 4 weeks in group B (n=8). T2-weighted images were obtained using an extremity coil. On MR imaging, we measured the signal intensity of the tumor and the uterine wall. We also measured the size and shape of the tumor and we compared this with the histopathologic results. RESULTS: On MR images obtained 2 weeks after inoculation, all the rabbit uteruses (group A and B, n=13) show a thick tubular wall, and the uteruses demonstrated a high signal intensity on T2-weighted images. The thickened uterine walls were measured as 3-10 mm (mean: 6.5 mm). Peritoneal nodules were observed in 3/13 rabbits (23%), the nodules measured 1.5-1.8x3.0 cm; uterine masses were observed in 3/13 rabbits (23%), and they measured 0.6-1.5x1.3-1.5 cm. On MR images obtained 4 weeks after the inoculations (group B, n=8), rabbit uteruses that had VX-II carcinoma show thick tubular wall in all cases (n=8, 100%) and round uterine masses (n=6, 75%). The thickened uterine walls measured 5-14 mm (mean 7.6 mm) and uterine masses measured 1.3-3.0x1.5-6.0 cm. All the rabbit uteruses having VX-II carcinoma showed high signal-to-noise ratios along the thickened uterine walls or masses on T2WI. On the histopathologic exam after sacrificing the rabbit, a few viable tumor cells were found because of necrosis in the inner portion of the uterine tumors, and abundant viable tumor cells were found at the periphery of the uterine tumors. CONCLUSION:We can develop an animal model with uterine tumor in rabbit uterus using VX-II carcinoma. Experimentally induced VX-II carcinomas in rabbit uterus demonstrated central necrosis, and MR imagings of experimentally induced uterine VX-II carcinomas were well correlated with the histopathology.

CT and MR Imaging Findings of Endometrial Stromal Sarcomas

PURPOSE: To evaluate the imaging findings of endometrial stromal sarcoma (ESS) according to histopathologic grade. MATERIALS AND METHODS: Six patients with pathologically proven ESS were included in this study. The histopathologic diagnosis was low-grade ESS for three patients and high-grade ESS for the three others. Preoperative CT or MR images were evaluated in terms of tumor size, location, growth pattern, the presence of hemorrhage or necrosis, status of the endometrial cavity, and invasion of surrounding structures. The imaging features of ESSs, which varied according to their histopathologic grade, were compared. RESULTS: The mean maximal diameter of low-and high-grade ESSs was 6 cm and 11.2 cm, respectively. All three low-grade ESSs were located mainly in the myometrium, but two high-grade ESSs were situated in the endometrial cavity and associated with focal tumor extension into the myometrium. One high-grade ESS had completely replaced the uterus. Low-grade ESSs were relatively well-defined, but high-grade ESSs had an irregular and lobulated margin. Intratumoral hemorrhage and necrosis were, respectively, found in two and three high-grade ESSs. Widening of the endometrial cavity was noted in all three high-grade ESSs, and lymph node metastasis had occurred in one. CONCLUSION: The imaging findings of ESS vary from a well-defined intramural mass to a bulky infiltrating mass, and depend on their histopathologic grade.

Cervical Carcinoma vs Endometrial Carcinoma, Involving Both Corpus and Cervix: Comparison of Growing Pattern with MR Imaging

PURPOSE: To evaluate the growth pattern depicted by MR imaging and used to differentiate between uterine cervcal and endometrial carcinoma where the mass involves both the uterine corpus and cervix. MATERIALS AND METHODS: The tumor growth pattern observed on MR images obtained between November 1989 and January in 1999 in 37 of 784 cervical carcinomas and 9 of 47 endometrial carcinomas in which the tumor involved both the uterine corpus and cervix was analysed. The histologic type was squamous (n=29), adenocarcinomatous (n=6) or adenosquamous (n=2) in cervical carcinoma, and carcinomatous (n=8) or adenosquamous (n=1) in endometrial carcinoma. A 1.5-T (Magnetom Vision, Siemens, Germany) and a 2.0-T unit (Spectro-20000, Goldstar, Korea) were used to obtain T1-and T2-weighted axial, T2-weighted sagittal and Gdenhanced images. Tumor involvement of the uterine cervix was classified as either partial(Cp) or total(Ct), and partial involvement(Cp) was subclassified as Cp-n, Cp-x, or Cp-b according to involvement of the endocervix, exocervix or both. Tumors of the uterine corpus were classified as involving the mucosa(U-mu), myometrium(U-my) or serosa(U-se). RESULTS: In 37 cases of cervical carcinoma, all three involving the endocervix(Cp-n) invaded the endometrium(U-mu), three involving both the endo- and exocervix(Cp-b) invaded the endometrium(U-mu, 1 case), myometrium(U-my, 1 case), or serosa(U-se, 1 case), and 31 involving the full-thickness of the uterine cervix(Ct) invaded the endometrium (U-mu, 6 cases) or serosa(U-se, 25 cases). In nine cases of endometrial carcinoma, three involving the endometrium(U-mu) and five involving the myometrium(U-my) invaded the endocervix(Cp-n), and one involving the serosa(U-se) invaded the full-thickness of the uterine cervix(Ct). CONCLUSION: Cervical carcinoma tended to involve the entire cervix and the full thickness of the uterine corpus, but endometrial carcinoma tended to involve the endometrium or myometrium of the uterine corpus and endocervix.

Recurrent Uterine Cervical Carcinoma:Spectrum of Imaging Findings

Uterine cervical carcinoma is one of the most common malignant tumors occur-ring in females. After primary treatment, patients are usually followed up with CT or MRI and the findings of these modalities may be the first sign of recurrent disease. Because earlier additional treatment by chemotherapy or radiation therapy may improve the prognosis, the early detection of recurrent cervical carcinoma is clinically important. In this article, we review the CT and MR imaging findings of recurrent uterine cervical carcinoma, and assign them to one of four groups: a)recurrence at the primary site, involving the intrapelvic organs, b) extension to the pelvic side-wall, c) metastases to pelvic and extrapelvic lymph nodes, or d)metastases to distant organs. A further contribution of CT and MR imaging is the detection of hydronephrosis due to ureteral obstruction. The cases in each group are illustrated and discussed, and since an awareness of the spectrum of imaging findings of recurrent cervical carcinoma is likely to lead to its early detection, radi-ologists should be familiar with the information presented.

Uterine Lipoma: A Case Report

A lipomatous tumor originating in the uterus is a rare benign neoplasm, and most reported cases have been of the mixed type. Authenticated cases of pure lipomas of the uterus are extremely rare. We report one such case in which the findings of magnetic resonance imaging provided the basis for preoperative diagnosis.

T2-weighted fast spin-echo MR findings of adenocarcinoma of the uterine cervix: comparison with squamous cell carcinoma

The purpose of this study was to investigate the differences in MR findings of adenocarcinoma (AC) and squamous cell carcinoma (SCC) of the uterine cervix and to compare MR findings with pathologic findings. MR images of 17 patients with pathologically proven AC, using a fast spin-echo (FSE) T2-weighted image (T2WI) with pelvic phased-array coil on a 1.5-T unit, were retrospectively evaluated. After measurement of the signal intensity (SI) ratios of the region of interest between tumors and gluteus maximus muscle, we compared the ratios of AC with those of 16 patients with SCC. AC showed relatively high SI on FSE T2WI with multiseptated lesions in four cases and hydrometrocolpos in three cases. The mean SI ratio was 3.82 +/- 1.68 in AC and 2.35 +/- 0.42 in SCC (p < 0.0001, t-test). Multiple tumorous glands with cytoplasmic and intraglandular mucin or serous fluid were pathologically found in AC, but SCC revealed the compact cellularity of stratified squamous tumor cells. The cervical AC showed higher SI than SCC on FSE T2WI with occasional multiseptated lesions and hydrometrocolpos. If the SI ratio of the tumor was more than 3.0, AC could be diagnosed with a sensitivity of 68.8% and a specificity of 100%.

CT and MR Findings of Placental Site Trophoblastic Tumor: A Report of Two Cases

Placental site trophoblastic tumor(PSTT) is a very rare form of gestational trophoblastic disease; weencountered two cases. MR imaging of the first patient revealed a solid uterine mass, slightly hyperintense tonormal myometrium on T1-weighted images and hyperintense on T2-weighted images. After intravenous gadoliniuminfusion, the mass was more enhanced than normal myometrium, and PSTT was confirmed pathologically after totalhysterectomy. In the second patient, PSTT was proven pathologically after laparoscopic left adnexectomy and wedgeresection of the right ovary. Postoperative enhanced CT scans revealed a well-enhanced right ovarian mass andextensive metastasis to the liver, pancreas, both adrenals, the stomach, lung and brain, but no mass was detectedin the uterus. Many of these metastatic lesions were infiltrating low density masses associated with highlyenhanced tubular or serpiginous structures, while some were of very high density.

Stage 1 Carcinoma of Cervix: Value of Endovaginal Surface Coil MR Imaging in Determining Depth of Invasion

PURPOSE: The purpose of this study was to determine the usefulness, with respect to depth of invasion, ofendovaginal surface coil (EVC) MR imaging for the evaluation of early-stage cervical cancer. MATERIALS AND METHODS: We retrospectively analyzed the MR images of 49 patients with pathologically-confirmed stage one cervicalcarcinoma who underwent EVC MR before hysterectomy. The tumor seen on MR images was identified as a high signalintensity lesion replacine the normal low signal intensity cervical stroma seen on T2WI. Depth of invasion, asseen on MR images and in pathologic specimens was divided into3-5mm, and the results werecompared. RESULTS: Depth of invasion in pathologic specimens was 5mm in25 patients(mean 6.34mm). On MR imaging, tumor was detected in 42 patients. Depth of invasion on MR was 5mm in 25 patients (mean 6.59mm). MR failed to detect tumors which infiltratedless than a certain depth;this cut-off point was 1mm in two patients, 2mm in two, and 3.5mm in one. All tumorswhich invaded more than 5mm were detected on EVC MR. There was no significant difference between depth of invasion seen on MR and in pathologic specimens(p=.59). The 95% confidence interval for difference in the value ofdepth of invasion between MR and on pathologic specimens ranged from -0.7mm to 1.2mm. CONCLUSION: EVC MR imagingmay be useful for assessing the depth of invasion of stage one cervical carcinoma.

CT and MRI Findings of Metastatic Uterine Carcinoma from Gastric Cancer: A Case Report

Uterine metastasis from stomach carcinoma occurred in a patient who five years previously had undergonegastrectomy for gastric adenocarcinoma. CT scans showed an ill-defined lobulated low density mass in the center ofthe uterus. T1-weighted MRI imaging showed that the lesion was indistinct from the myometrium; on T2-weightedimaging it showed increased signal intensity and had infiltrated the myometrium. After Gd-DTPA administration, thelesion showed very little enhancement.

'Bridging Vascular Sign' in the MR Diagnosis of Uterine Subserosal/Exophytic Leiomyoma

PURPOSE: To evaluate the usefulness of 'bridging vascular sign' in the diagnosis of a subserosal/exophyticuterine leiomyoma and the helpfulness of this sign in the differentiation of a uterine leiomyoma from adnexalmasses on pelvic magnetic resonance (MR) imaging. MATERIALS AND METHODS: Of 20 women with a solitary pelvic massin whom pelvic MR imaging was performed during recent three years to differentiate a uterine leiomyoma from anadnexal mass, 12 with surgicopathologically proven subserosal/exophytic uterine leiomymas were included in thisstudy. The other eight women were proved to have tubo-ovarian abscess (n = 3), ovarian endometrioma (n = 2),ruptured corpus luteum cyst (n = 2), and ovarian fibroma (n = 1). A 1.5 T unit was used to obtain axial/sagittalT1- and T2-weighted and gadolinium-enhanced sagittal T1-weighted images. Positive 'bridging vascular sign' wasdefined through the retrospective review and analysis of MR findings as the presence of intervening curvilinear ortortuous signal-void vascular structures crossing the uterus and the pelvic mass. Using this sign in the diagnosisand differentiation of a uterine leiomyoma from adnexal masses, statistical significance was evaluated. RESULTS:'Bridging vascular sign' was present in eight of 12 leiomyomas, but no in adnexal masses. Using this sign in thediagnosis and differentiation of a uterine leiomyoma from adnexal masses, sensitivity was 66.7%, specificity 100%,positive predictive value 100%, negative predictive value 66.7%, and accuracy 80%. CONCLUSION: 'Vascular bridgingsign' on MR imaging may be a useful radiologic sign in the diagnosis and differentiation of a subserosal/exophyticuterine leiomyoma from adnexal masses.

Carcinoma of the Cervix: Usefulness of Dynamic and Contrast-Enhanced T1-weighted MR Imaging in Assessing theDepth of Stromal Invasion

PURPOSE: The purpose of this study was to evaluate the usefulness of contrast enhancement in assessing thedepth of stromal invasion in patients with uterine cervical tumors by comparing dynamic and late contrast-enhancedT1-weighted MR imaging with T2-weighted MR imaging. MATERIALS AND METHODS: Of 58 surgically proven uterinecervical cancer patients, 31 in whom tumors were seen on MRI were included in this study. Using a 1.5 T magnet,T2-weighted contrast-enhanced dynamic, and T1-weighted MR imaging were performed. In each MR imaging sequence,tumor visualization, margin delineation, enhancement pattern and depth of stromal invasion were evaluated onsagittal images and were correlated with pathological findings of resected uterus, focusing on the depth ofstromal invasion. RESULTS: Surgical FIGO stages were IB1 in 20 patients, IB2 in three, IIA in six, and IIB intwo. A tumor was detected in 29 (94%) patients on T2-weighted images, in 26 (84%) on dynamic contrast enhancedimages, and in 28 (90%) on contrast enhanced T1-weighted images. The tumor demonstrated a clear margin in 22 (71%)patients on late contrast-enhanced T1-weighted images, in 21 (68%) on dynamic images, and in 13 (42%) onT2-weighted images. Correlated with pathologic findings, the depth of stromal invasion was overestimated onT2-weighted images in eight (26%) patients, and on dynamic and late contrast-enhanced T1-weighted images in three(10%) and three (10%) respectively. CONCLUSION: Tumor margins are clearer and the extent of tumors may be moreaccurately evaluated on dynamic MR and late contrast-enhanced T1-weighted imaging. These sequences thas seem to beuseful, and superior to T2-weighted imaging, for assessing the depth of stromal invasion in patients with cervicalcarcinoma.

MR Imaging of Uterine Cervical Carcinoma: Is Contrast-enhanced Imaging Useful in Staging?

PURPOSE: To evaluate the usefulness of contrast-enhanced T1-weighted MR imaging in the diagnosis and stagingof uterine cervical carcinoma. MATERIALS AND METHODS: MR images (T1WI, T2WI and CE-T1WI) of 40 patients withhistologically proven uterine cervical carcinoma were reviewed by three radiologists. Tumor visualization, signalintensity, degree of enhancement, delineation of tumor margin, assessment of parametrial involvement, andevaluation of pelvic lymph node metastasis were analynized on each sequence, retrospectively. We compared theaccuracy of CE-T1WI with that of T2WI in the diagnosis and staging of uterine cervical carcinoma. RESULTS: Tumorswere detected in 35/40 cases (87.5%) on T2WI, and in 34/40 (85.0%) on CE-T1WI. On T2WI, signal intensities of allvisualized tumors were more hyperintense than those of normal cervical stroma. The signal intensities of enhancedtumors were lower(79.4%), similar to (5.9%), or higher than (14.7%) those of well-enhanced normal myometrium. Inthe assessment of parametrial involvement and the evaluation of pelvic lymph node metastasis, accuracy during eachsequence was 85.7% (12/14) and 100% (28/28), respectively. With regard to delineation of tumor margin, nostatistical difference was found between CE-T1WI and T2WI, though in cases of lesions greater than 3 cm in maximalsize, the margin was clearer in 14 cases, and less clear in 4 cases, on CE-T1WI than on T2WI, with a significantstatistical difference. The accuracy of tumor staging was 78.6% (11/14) on T2WI, and 71.4% (10/14) on CE-T1WI. CONCLUSION: There was no significant statistical difference in tumor visualization, accuracy of staging, andevaluation of pelvic lymph node metastasis between T2W1 and CE-T1WI. CE-T1WI was superior to T2WI in delineatingthe margin of tumors more than 3 cm in maximal size. Contrast-enhanced T1WI appears to be useful for evaluatingthe extent of large cervical carcinomas.

Undetectable, Small Uterine Cervical Tumors on MR Imaging: Comparison to Detectable Tumors

PURPOSE: Small uterine cervical tumors are often undetected on MRI, and undetectable tumors are usually regarded as being in their early stages. The purpose of this study was to evaluate morphological factors determining MRI detectability of uterine cervical tumors by comparing detected and undetected tumors using high resolution MRI units. MATERIALS AND METHODS: Sixty-six surgically-proven uterine cervical cancer patients were included in this study. Using a GE Signa 1.5T magnet (GE, U.S.A.) axial T1-weighted MR images and axial, sagittal, and coronal T2-weighted fast spin echo MR images were obtained with a 5 mm thickness/2 mm gap and a 512x256 matrix size. The patients were divided into two groups, tumor-detected and tumor-undetected. Maximal tumor dimension, depth of stromal invasion, and horizontal tumor spread, measured during histopathological examinations, as well as pathological stages, were compared between the two groups. RESULTS: In the tumor-detected group, the pathological stages were IB1 in 25, IB2 in 3, IIA in 10, and IIB in 5 patients; in the tumor-undetected group, the stages were IA1 in 8, IA2 in 2, and IB1 in 13 patients. Maximal tumor dimension, depth of stromal invasion and horizontal tumor spread for tumor-detected and -undetected groups were 27.4 (10-60) mm vs. 10.4 (3-40) mm (p=0.077), 11.1 (3-20) mm vs. 2.3 (0-5) mm (p<0.001), and 26.6 (8-60) mm vs. 10.4 (3-40) mm (p=0.057), respectively. CONCLUSION: Of the criteria studied, depth of stromal invasion is the most important factor in determining tumor detectability on MRI. All tumors in which this depth was more than 5 mm were thus detected; which are undetected on high resolution MRI can be regarded as early-stage tumors (stage IB1).

MR Imaging of Gestational Trophoblastic Tumor: Role of Gadolinium Enhancement

PURPOSE: To investigate the role of gadolinium enhanced MR imaging in the evaluation of gestational trophoblastic tumors (invasive mole and choriocarcinoma). MATERIALS AND METHODS: Pre-enhanced T1-and T2-weighted images and gadolinium enhanced T1-weighted images of 34 gestational trophoblastic tumors (15 choriocarcinomas, 19 invasive moles) were evaluated retrospectively. Enhanced patterns of gestational trophoblastic tumors were analyzed. Morphologic differences and structural characterstics were analyzed by the evaluation of tumor margin, patterns of hemorrhagic necroses, development of intratumoral vascularity, and molar villi. Graded scores of MR findings between pre- and gadolinium enhanced images were given in the following criteria ; 1) visualization of tumor margin 2) distinction between tumor necrosis and zone of trophoblastic proliferation, and 3) molar villi. Statistical differences between graded scores of pre-and post-enhanced images were analyzed. RESULTS: Choriocarcinoma was a well-defined mass with peripheral rim enhancement and central hemorrhagic necrosis. Invasive mole was a ill-defined mass with partial necrosis, tiny cystic areas and increased intratumoral vascularity with dense reticular enhancement. Interface between tumor necrosis and zone of trophoblastic proliferation was better visualized on the gadolinium enhanced images than on the pre-enhanced images. Tumor margin and intratumoral molar villi were clearly discrimated on the gadolinium enhanced images only in invasive mole. There was no statistical difference in the visualization of tumor margin of choriocarcinoma between the pre-and post-enhanced images. CONCLUSION: Gadolinium enhanced MR image was helpful in the visualization of tumor characterstics in gestational trophoblastic tumors, and in the differential diagnosis between invasive mole and choriocarcinoma.

Locally Advanced Squamous Cell Carcinoma of the Uterine Cervix: MR Evaluation of Effect of Combined Biologic (13-cis-Retinoic Acid plus Interferon alpha-2a) and Radiation Therapy

PURPOSE: To evaluate the effectiveness of MR imaging in assessing the response of carcinomas of the uterinecervix to the combined therapy of interferon alpha-2a, 13-cis-retinoic acid and radiation. MATERIALS AND METHODS: Thirteen patients with untreated, locally advanced squamous cell cervical carcinoma were treated for eight weeks with 13-cis-retinoic acid plus interferon alpha-2a and radiotherapy. Axial and sagittal T1-weighted spin echo and fast spin echo MR images were obtained before and after treatment, using a 1.5-T MR scanner. MR images were correlated with biopsy findings. In each patient, pre- and post-treatment images were prospectively analyzed andcompared. RESULTS: In all patients, tumor volume could be calculated by three dimensional measurement on MR images. Initial response to the combined therapy was complete in eight patients, partial in four and progressivein one. Evidence of tumor recurrence was noted in two patients during the follow-up period. MR images correlated well with biopsy findings. CONCLUSION: MR imaging is useful in the evaluation of tumor response to the combined biologic and radiation therapy, which can be an effective first-line therapy against locally advanced squamouscell carcinoma of the cervix.

Endometrial Carcinoma Occuring from Polycystic Ovary Disease: A Case Report

Endometrial carcinoma usually occurs in postmenopausal women ; less than 5% occurs in women under the age of40. Up to one quarter of endometrial carcinoma patients below this age have PCO(polycystic ovary disease, Stein-Leventhal syndrome). The increased incidence of endometrial carcinoma in patients with PCO is related to chronic estrogenic stimulation. We report MR imaging in one case of endometrial carcinoma occuring in a 23 yearold woman with PCO and had complained of hypermenorrhea for about three years. On T2-weighted MR image the endometrial cavity was seen to be distended with protruded endometrial masses of intermediate signal intensity, and the junctional zone was disrupted beneath the masses. Both ovaries were best seen on T2-weighted MR imagingand showed multiple small peripheral cysts and low signal-intensity central stroma.